6.4 Crossing the midline: a case study

Many neurons on the left side of the body make contact with targets on the right, contralateral, side of the body, and vice verse. Crossing the midline is particularly prevalent in descending neurons (e.g. the Betz cells mentioned earlier) and interneurons in the spinal cord which form part of the sensory pathways. However, not all spinal cord interneurons cross the midline: short distance interneurons which influence motor neurons and certain neurons of the spinoreticular tract are ipsilateral. We now look at how the growing axon is guided to cross the midline and become a commissural axon or to remain ipsilateral and become a longitudinal axon.

In 1993 Seeger reported two Drosophila mutants in which axon growth across the midline was affected. In the first, called roundabout or robo, axons crossed the midline normally, but then re-crossed and crossed again. robo mutants have axons which cross the midline freely. (There is a written convention here: in Drosophila, the word in italics, e.g. robo, refers to the mutant organism or mutant gene, which does not produce a functioning gene product; the word in normal script, robo, refers to a functioning protein, the gene product.)

The second mutant had axons which never crossed the midline. This mutant is called commissureless or comm. comm mutants have axons which cannot cross the midline. (See Figures 17 and 18.)

A fly with both these mutations has axons which cross the midline freely.

Figure 17 Diagramatic representation of longitudinal and commissural axons. (a) In the normal situation longitudinal axons remain ipsilateral, commissural axons cross the midline. (b) In comm mutants very few axons cross the midline. (c) In robo mutants most axons can cross the midline resulting in enlarged commissural pathways. (d) In sli mutants all axons can enter the midline but cannot then leave it, resulting in both pronounced commissures and barely visible longitudinal pathways

View larger image

Figure 18 The photographs from which the drawings in Figure 17 are taken. (a) Normal, wild type, (b) comm, (c) robo, (d) sli

These data can be explained in numerous ways, which is often the case in developmental studies, but a satisfactory explanation had to wait until 1999 when Kidd discovered an important repellent molecule in the extracellular matrix of the midline (Kidd et al., 1999). The molecule is called sli and it repels most growth cones.

sli and robo interact. sli is the key, or ligand which interacts with the receptor robo.

For the growth cone to cross the midline, then, one of two things must happen.

If the midline cells stop producing sli then any growth cones could cross the midline. As only some growth cones cross the midline, control must be exerted through the growth cones themselves; thus it is the growth cones that stop producing robo that are able to cross the midline.

Two problems remain. The first is, what is to stop the axons that do cross the midline from crossing it again, as the robo mutants do? The second is where does comm fit in?

The answer to the first question is that once across the midline, the growth cones of commissural axons begin to produce robo. The mechanism that switches on robo production at the right moment has not been worked out.

The answer to the second question is that comm is an intracellular signal that inhibits (down regulates) the production of robo; in the presence of comm, little or no robo is produced. When comm is not produced (i.e. in comm mutants) robo is produced and no growth cones can cross the midline.

There is a parallel medical condition in humans in which cortical axons fail to cross the midline to connect with neurons in the cortex on the opposite side of the brain. The condition, agenesis (pronounced ‘a-genesis’) of the corpus callosum, is not well understood, but may well result from overproduction of a protein very similar to robo.